We propose a study to examine the HIV/AIDS risk behaviors of homeless youth in large interconnected networks of youth. The nearly 2 million runaway and homeless youth (RHY) in the United States each year are at great risk for contracting HIV/AIDS, with prevalence rates reported as high as 11.5%. To develop new interventions or improve existing network-based interventions, we must have an accurate assessment of network structure. The requisite data have not been collected because RHY are an unbounded population, and social network researchers have not reached consensus as to how to collect sociometric data from unbounded populations. The event-based approach (EBA) proposed by Freeman that samples the regulars of a socio-physical space (e.g. a beach) seems promising for RHY and will be tested in the proposed study. A broad spectrum of RHY (those in shelters and on the streets) can be found in large numbers, habitually accessing services from drop-in centers where they socialize, eat, and receive case management - they are regulars of a socio-physical space. These centers are also sustainable sites for future intervention delivery. Thus our Specific Aims are: 1. To use the event-based approach (EBA) to collect multiple panels of sociometric network data over time on an unbounded population of youth at risk for HIV/AIDS. 2. To examine how network positions relate to risk-taking and whether risk-taking structures endure over time, independent of the individual youth who may occupy those positions in the network. 3. To examine network ties to pro-social home-based peers and supportive family (including online, cell phone, and face to face connections) to explore sources of healthy influence and new points of intervention for youth who are not embedded in larger RHY networks. To accomplish these aims, we propose a three year study. RHY network data will be collected in two drop-in centers serving RHY in different neighborhoods in Los Angeles. We will collect four panels of sociometric data, once every 6 months at each location. Successfully addressing these aims will provide: (1) unique over time sociometric data on HIV risk in unbounded networks, (2) direction as to how to spend prevention dollars targeting RHY with face-to-face network interventions, and (3) new directions in HIV prevention interventions, using social networking technology to connect RHY to pro-social, non-RHY networks.